# CS 506 Fall 2025 Technical Midterm — Data

来源：https://www.kaggle.com/competitions/cs-506-fall-2025-technical-midterm/data

采集时间：2026-10-19T08:53:50Z

## Files

*   **train.csv** - Contains 447,583 unique reviews. 80,000 of these reviews will have a missing score column. These are the scores you want to predict
*   **test.csv** - A two-column table, where the first column is the 80k reviews mentioned above with a missing score column, and the second column is the Score column you are intended to make predictions with. You will use train.csv to make these predictions
*   **sample_submission.csv** - a sample submission file in the correct format

## Columns

*   `id` - Unique identifier for every individual review
*   `reviewerID` - Unique identifier for the reviewer who wrote the review
*   `album_mbid` - Unique identifier for the specific album being reviewed
*   `artist_mbid` - Unique identifier for the artist that made the album being reviewed
*   `unixReviewTime` - The time the review was made, described in Unix Time
*   `VotedHelpful` - The number of people who voted on a review, and marked it as helpful
*   `TotalVotes` - The total number of people who voted on the review, whether voted helpful or not
*   `summary` - A brief summary of the review
*   `reviewText` - The pure text content of the review
*   `genres` - A comma-separated list of the genres the album could be classified under
*   `Score` - An integer rating in between 1 and 5. Blank for the fields you are intended to predict

Note: Some of the rows of the table may have some of these values missing.